diff --git a/backends/advanced/init.py b/backends/advanced/init.py
index 8d2e51f6..4e345ec1 100644
--- a/backends/advanced/init.py
+++ b/backends/advanced/init.py
@@ -15,7 +15,7 @@
 from pathlib import Path
 from typing import Any, Dict
 
-from dotenv import set_key
+from dotenv import get_key, set_key
 from rich.console import Console
 from rich.panel import Panel
 from rich.prompt import Confirm, Prompt
@@ -100,6 +100,26 @@ def backup_existing_env(self):
             shutil.copy2(env_path, backup_path)
             self.console.print(f"[blue][INFO][/blue] Backed up existing .env file to {backup_path}")
 
+    def read_existing_env_value(self, key: str) -> str:
+        """Read a value from existing .env file"""
+        env_path = Path(".env")
+        if not env_path.exists():
+            return None
+
+        value = get_key(str(env_path), key)
+        # get_key returns None if key doesn't exist or value is empty
+        return value if value else None
+
+    def mask_api_key(self, key: str, show_chars: int = 5) -> str:
+        """Mask API key showing only first and last few characters"""
+        if not key or len(key) <= show_chars * 2:
+            return key
+
+        # Remove quotes if present
+        key_clean = key.strip("'\"")
+
+        return f"{key_clean[:show_chars]}{'*' * min(15, len(key_clean) - show_chars * 2)}{key_clean[-show_chars:]}"
+
     def setup_authentication(self):
         """Configure authentication settings"""
         self.print_section("Authentication Setup")
@@ -128,8 +148,17 @@ def setup_transcription(self):
         if choice == "1":
             self.console.print("[blue][INFO][/blue] Deepgram selected")
             self.console.print("Get your API key from: https://console.deepgram.com/")
-            
-            api_key = self.prompt_value("Deepgram API key (leave empty to skip)", "")
+
+            # Check for existing API key
+            existing_key = self.read_existing_env_value("DEEPGRAM_API_KEY")
+            if existing_key and existing_key not in ['your_deepgram_api_key_here', 'your-deepgram-key-here']:
+                masked_key = self.mask_api_key(existing_key)
+                prompt_text = f"Deepgram API key ({masked_key}) [press Enter to reuse, or enter new]"
+                api_key_input = self.prompt_value(prompt_text, "")
+                api_key = api_key_input if api_key_input else existing_key
+            else:
+                api_key = self.prompt_value("Deepgram API key (leave empty to skip)", "")
+
             if api_key:
                 self.config["TRANSCRIPTION_PROVIDER"] = "deepgram"
                 self.config["DEEPGRAM_API_KEY"] = api_key
@@ -141,10 +170,19 @@ def setup_transcription(self):
             self.config["TRANSCRIPTION_PROVIDER"] = "mistral"
             self.console.print("[blue][INFO][/blue] Mistral selected")
             self.console.print("Get your API key from: https://console.mistral.ai/")
-            
-            api_key = self.prompt_value("Mistral API key (leave empty to skip)", "")
+
+            # Check for existing API key
+            existing_key = self.read_existing_env_value("MISTRAL_API_KEY")
+            if existing_key and existing_key not in ['your_mistral_api_key_here', 'your-mistral-key-here']:
+                masked_key = self.mask_api_key(existing_key)
+                prompt_text = f"Mistral API key ({masked_key}) [press Enter to reuse, or enter new]"
+                api_key_input = self.prompt_value(prompt_text, "")
+                api_key = api_key_input if api_key_input else existing_key
+            else:
+                api_key = self.prompt_value("Mistral API key (leave empty to skip)", "")
+
             model = self.prompt_value("Mistral model", "voxtral-mini-2507")
-            
+
             if api_key:
                 self.config["MISTRAL_API_KEY"] = api_key
                 self.config["MISTRAL_MODEL"] = model
@@ -178,11 +216,20 @@ def setup_llm(self):
             self.config["LLM_PROVIDER"] = "openai"
             self.console.print("[blue][INFO][/blue] OpenAI selected")
             self.console.print("Get your API key from: https://platform.openai.com/api-keys")
-            
-            api_key = self.prompt_value("OpenAI API key (leave empty to skip)", "")
+
+            # Check for existing API key
+            existing_key = self.read_existing_env_value("OPENAI_API_KEY")
+            if existing_key and existing_key not in ['your_openai_api_key_here', 'your-openai-key-here']:
+                masked_key = self.mask_api_key(existing_key)
+                prompt_text = f"OpenAI API key ({masked_key}) [press Enter to reuse, or enter new]"
+                api_key_input = self.prompt_value(prompt_text, "")
+                api_key = api_key_input if api_key_input else existing_key
+            else:
+                api_key = self.prompt_value("OpenAI API key (leave empty to skip)", "")
+
             model = self.prompt_value("OpenAI model", "gpt-4o-mini")
             base_url = self.prompt_value("OpenAI base URL (for proxies/compatible APIs)", "https://api.openai.com/v1")
-            
+
             if api_key:
                 self.config["OPENAI_API_KEY"] = api_key
                 self.config["OPENAI_MODEL"] = model
diff --git a/backends/advanced/ssl/generate-ssl.sh b/backends/advanced/ssl/generate-ssl.sh
index fb83db5b..efc5d8c2 100755
--- a/backends/advanced/ssl/generate-ssl.sh
+++ b/backends/advanced/ssl/generate-ssl.sh
@@ -2,24 +2,26 @@
 set -e
 
 # Generate self-signed SSL certificate for Friend-Lite Advanced Backend
-# Supports localhost and custom Tailscale IP
+# Supports localhost, IP addresses, and domain names
 
-TAILSCALE_IP="$1"
+SERVER_ADDRESS="$1"
 
-if [ -z "$TAILSCALE_IP" ]; then
-    echo "Usage: $0 <tailscale-ip>"
+if [ -z "$SERVER_ADDRESS" ]; then
+    echo "Usage: $0 <ip-or-domain>"
     echo "Example: $0 100.83.66.30"
+    echo "Example: $0 myserver.tailxxxxx.ts.net"
     exit 1
 fi
 
-# Validate IP format
-if ! echo "$TAILSCALE_IP" | grep -E '^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$' > /dev/null; then
-    echo "Error: Invalid IP format. Expected format: xxx.xxx.xxx.xxx"
-    exit 1
+# Detect if it's an IP address or domain name
+if echo "$SERVER_ADDRESS" | grep -E '^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$' > /dev/null; then
+    IS_IP=true
+    echo "🔐 Generating SSL certificate for localhost and IP: $SERVER_ADDRESS"
+else
+    IS_IP=false
+    echo "🔐 Generating SSL certificate for localhost and domain: $SERVER_ADDRESS"
 fi
 
-echo "🔐 Generating SSL certificate for localhost and $TAILSCALE_IP"
-
 # Determine the output directory - we should be in backends/advanced when running
 SSL_DIR="ssl"
 if [ -d "$SSL_DIR" ]; then
@@ -53,9 +55,15 @@ subjectAltName = @alt_names
 DNS.1 = localhost
 DNS.2 = *.localhost
 IP.1 = 127.0.0.1
-IP.2 = $TAILSCALE_IP
 EOF
 
+# Add custom address as either IP or DNS
+if [ "$IS_IP" = true ]; then
+    echo "IP.2 = $SERVER_ADDRESS" >> $OUTPUT_DIR/server.conf
+else
+    echo "DNS.3 = $SERVER_ADDRESS" >> $OUTPUT_DIR/server.conf
+fi
+
 # Generate private key
 openssl genrsa -out $OUTPUT_DIR/server.key 2048
 
@@ -75,7 +83,7 @@ chmod 644 $OUTPUT_DIR/server.crt
 echo "✅ SSL certificate generated successfully"
 echo "   - Certificate: $OUTPUT_DIR/server.crt"
 echo "   - Private key: $OUTPUT_DIR/server.key"
-echo "   - Valid for: localhost, *.localhost, 127.0.0.1, $TAILSCALE_IP"
+echo "   - Valid for: localhost, *.localhost, 127.0.0.1, $SERVER_ADDRESS"
 echo ""
 echo "Certificate Details:"
 openssl x509 -in $OUTPUT_DIR/server.crt -text -noout | grep -A 1 "Subject Alternative Name" || echo "   (Certificate generated successfully)"
\ No newline at end of file
diff --git a/extras/local-omi-bt/start.sh b/extras/local-omi-bt/start.sh
index c59c44e2..6fd8947e 100755
--- a/extras/local-omi-bt/start.sh
+++ b/extras/local-omi-bt/start.sh
@@ -1,2 +1,2 @@
 #!/bin/bash
-uv run --with-requirements requirements.txt python connect-laptop.py
+uv run --with-requirements requirements.txt python connect-omi.py
diff --git a/extras/speaker-recognition/.env.template b/extras/speaker-recognition/.env.template
index bc264b44..2ba3a1c7 100644
--- a/extras/speaker-recognition/.env.template
+++ b/extras/speaker-recognition/.env.template
@@ -13,6 +13,11 @@ HF_TOKEN=your_huggingface_token_here
 # This controls which dependency group is installed in the Docker image
 COMPUTE_MODE=cpu
 
+# PyTorch CUDA version (only used when COMPUTE_MODE=gpu)
+# Options: cu121 (CUDA 12.1), cu126 (CUDA 12.6), cu128 (CUDA 12.8)
+# Should match your system's CUDA version (check with: nvidia-smi)
+PYTORCH_CUDA_VERSION=cu126
+
 # Speaker recognition similarity threshold (0.0-1.0)
 # Lower values = less strict identification, higher values = more strict
 # Typical range: 0.1-0.3 for ECAPA-TDNN models
diff --git a/extras/speaker-recognition/Dockerfile b/extras/speaker-recognition/Dockerfile
index 40fccc5d..0fc84b84 100644
--- a/extras/speaker-recognition/Dockerfile
+++ b/extras/speaker-recognition/Dockerfile
@@ -1,9 +1,9 @@
 FROM python:3.12-bookworm
 
-ARG CUDA_VERSION
+ARG PYTORCH_CUDA_VERSION
 
 # Capture build arg as environment variable for runtime
-ENV CUDA_VERSION=${CUDA_VERSION}
+ENV PYTORCH_CUDA_VERSION=${PYTORCH_CUDA_VERSION}
 
 # Install system dependencies including PortAudio for pyaudio
 RUN apt-get update && apt-get install -y \
@@ -27,8 +27,8 @@ COPY pyproject.toml ./
 RUN mkdir -p src/simple_speaker_recognition
 COPY src/simple_speaker_recognition/__init__.py src/simple_speaker_recognition/
 
-# Install dependencies and package  
-RUN uv sync --no-dev --extra ${CUDA_VERSION}
+# Install dependencies and package
+RUN uv sync --no-dev --extra ${PYTORCH_CUDA_VERSION}
 
 # Create directories
 RUN mkdir -p /app/audio_chunks /app/debug /app/data /models
@@ -42,4 +42,4 @@ EXPOSE 8085
 
 # Run the service
 # Use shell form to allow environment variable expansion
-CMD uv run --extra ${CUDA_VERSION} --no-dev simple-speaker-service 
\ No newline at end of file
+CMD uv run --extra ${PYTORCH_CUDA_VERSION} --no-dev simple-speaker-service 
\ No newline at end of file
diff --git a/extras/speaker-recognition/docker-compose-test.yml b/extras/speaker-recognition/docker-compose-test.yml
index a4b13036..0cc6d8d2 100644
--- a/extras/speaker-recognition/docker-compose-test.yml
+++ b/extras/speaker-recognition/docker-compose-test.yml
@@ -5,7 +5,7 @@ services:
       context: .
       dockerfile: Dockerfile
       args:
-        CUDA_VERSION: ${COMPUTE_MODE:-cpu}
+        PYTORCH_CUDA_VERSION: ${COMPUTE_MODE:-cpu}
     image: speaker-recognition:test
     ports:
       # Map host test port (default 8086) to container port 8085
diff --git a/extras/speaker-recognition/docker-compose.yml b/extras/speaker-recognition/docker-compose.yml
index 52f2a80a..9118e732 100644
--- a/extras/speaker-recognition/docker-compose.yml
+++ b/extras/speaker-recognition/docker-compose.yml
@@ -8,7 +8,7 @@ services:
       context: .
       dockerfile: Dockerfile
       args:
-        CUDA_VERSION: ${CUDA_VERSION:-cpu}
+        PYTORCH_CUDA_VERSION: ${PYTORCH_CUDA_VERSION:-cpu}
     image: speaker-recognition:latest
     env_file:
       - .env
@@ -46,7 +46,7 @@ services:
       context: .
       dockerfile: Dockerfile
       args:
-        CUDA_VERSION: ${CUDA_VERSION:-cu121}
+        PYTORCH_CUDA_VERSION: ${PYTORCH_CUDA_VERSION:-cu121}
     deploy:
       resources:
         reservations:
@@ -108,3 +108,4 @@ services:
 networks:
   default:
     name: friend-network
+    external: true
diff --git a/extras/speaker-recognition/init.py b/extras/speaker-recognition/init.py
index 86797ed1..a5354394 100755
--- a/extras/speaker-recognition/init.py
+++ b/extras/speaker-recognition/init.py
@@ -65,6 +65,27 @@ def prompt_password(self, prompt: str) -> str:
                 self.console.print("[red][ERROR][/red] Token is required for speaker recognition")
                 sys.exit(1)
 
+    def read_existing_env_value(self, key: str) -> str:
+        """Read a value from existing .env file"""
+        env_path = Path(".env")
+        if not env_path.exists():
+            return None
+
+        from dotenv import get_key
+        value = get_key(str(env_path), key)
+        # get_key returns None if key doesn't exist or value is empty
+        return value if value else None
+
+    def mask_api_key(self, key: str, show_chars: int = 5) -> str:
+        """Mask API key showing only first and last few characters"""
+        if not key or len(key) <= show_chars * 2:
+            return key
+
+        # Remove quotes if present
+        key_clean = key.strip("'\"")
+
+        return f"{key_clean[:show_chars]}{'*' * min(15, len(key_clean) - show_chars * 2)}{key_clean[-show_chars:]}"
+
     def prompt_choice(self, prompt: str, choices: Dict[str, str], default: str = "1") -> str:
         """Prompt for a choice from options"""
         self.console.print(prompt)
@@ -103,9 +124,65 @@ def setup_hf_token(self):
             self.config["HF_TOKEN"] = self.args.hf_token
             self.console.print("[green][SUCCESS][/green] HF Token configured from command line")
         else:
-            hf_token = self.prompt_password("HF Token")
-            self.config["HF_TOKEN"] = hf_token
-            self.console.print("[green][SUCCESS][/green] HF Token configured")
+            # Check for existing token
+            existing_token = self.read_existing_env_value("HF_TOKEN")
+            if existing_token and existing_token not in ['your_huggingface_token_here', 'your-hf-token-here']:
+                masked_token = self.mask_api_key(existing_token)
+                self.console.print(f"[blue][INFO][/blue] Found existing token: {masked_token}")
+                try:
+                    reuse = Confirm.ask("Use existing HF Token?", default=True)
+                except EOFError:
+                    reuse = True
+
+                if reuse:
+                    self.config["HF_TOKEN"] = existing_token
+                    self.console.print("[green][SUCCESS][/green] HF Token configured (reused)")
+                else:
+                    hf_token = self.prompt_password("HF Token")
+                    self.config["HF_TOKEN"] = hf_token
+                    self.console.print("[green][SUCCESS][/green] HF Token configured")
+            else:
+                hf_token = self.prompt_password("HF Token")
+                self.config["HF_TOKEN"] = hf_token
+                self.console.print("[green][SUCCESS][/green] HF Token configured")
+
+    def detect_cuda_version(self) -> str:
+        """Detect system CUDA version from nvidia-smi"""
+        try:
+            result = subprocess.run(
+                ["nvidia-smi", "--query-gpu=driver_version", "--format=csv,noheader"],
+                capture_output=True,
+                text=True,
+                timeout=5
+            )
+            if result.returncode == 0:
+                # Try to get CUDA version from nvidia-smi
+                result = subprocess.run(
+                    ["nvidia-smi"],
+                    capture_output=True,
+                    text=True,
+                    timeout=5
+                )
+                if result.returncode == 0:
+                    output = result.stdout
+                    # Parse CUDA Version from nvidia-smi output
+                    # Format: "CUDA Version: 12.6"
+                    import re
+                    match = re.search(r'CUDA Version:\s*(\d+)\.(\d+)', output)
+                    if match:
+                        major, minor = match.groups()
+                        cuda_ver = f"{major}.{minor}"
+
+                        # Map to available PyTorch CUDA versions
+                        if cuda_ver >= "12.8":
+                            return "cu128"
+                        elif cuda_ver >= "12.6":
+                            return "cu126"
+                        elif cuda_ver >= "12.1":
+                            return "cu121"
+        except (subprocess.SubprocessError, FileNotFoundError):
+            pass
+        return "cu121"  # Default fallback
 
     def setup_compute_mode(self):
         """Configure compute mode (CPU/GPU)"""
@@ -125,14 +202,44 @@ def setup_compute_mode(self):
 
         self.config["COMPUTE_MODE"] = compute_mode
 
-        # Set CUDA_VERSION for Docker build
+        # Set PYTORCH_CUDA_VERSION for Docker build
         if compute_mode == "cpu":
-            self.config["CUDA_VERSION"] = "cpu"
+            self.config["PYTORCH_CUDA_VERSION"] = "cpu"
         else:
-            # Default to cu121 for GPU mode (can be overridden in .env)
-            self.config["CUDA_VERSION"] = "cu121"
+            # Detect system CUDA version and suggest as default
+            detected_cuda = self.detect_cuda_version()
+
+            # Map to default choice number
+            cuda_to_choice = {
+                "cu121": "1",
+                "cu126": "2",
+                "cu128": "3"
+            }
+            default_choice = cuda_to_choice.get(detected_cuda, "2")
+
+            self.console.print()
+            self.console.print(f"[blue][INFO][/blue] Detected CUDA version: {detected_cuda}")
+            self.console.print()
+
+            cuda_choices = {
+                "1": "CUDA 12.1 (cu121)",
+                "2": "CUDA 12.6 (cu126)",
+                "3": "CUDA 12.8 (cu128)"
+            }
+            cuda_choice = self.prompt_choice(
+                "Choose CUDA version for PyTorch:",
+                cuda_choices,
+                default_choice
+            )
+
+            choice_to_cuda = {
+                "1": "cu121",
+                "2": "cu126",
+                "3": "cu128"
+            }
+            self.config["PYTORCH_CUDA_VERSION"] = choice_to_cuda[cuda_choice]
 
-        self.console.print(f"[blue][INFO][/blue] Using {compute_mode.upper()} mode")
+        self.console.print(f"[blue][INFO][/blue] Using {compute_mode.upper()} mode with PyTorch CUDA version: {self.config['PYTORCH_CUDA_VERSION']}")
 
     def setup_deepgram(self):
         """Configure Deepgram API key if provided"""
diff --git a/extras/speaker-recognition/ssl/generate-ssl.sh b/extras/speaker-recognition/ssl/generate-ssl.sh
index b76bf83a..c1e832c5 100755
--- a/extras/speaker-recognition/ssl/generate-ssl.sh
+++ b/extras/speaker-recognition/ssl/generate-ssl.sh
@@ -2,24 +2,26 @@
 set -e
 
 # Generate self-signed SSL certificate for Friend-Lite Advanced Backend
-# Supports localhost and custom Tailscale IP
+# Supports localhost, IP addresses, and domain names
 
-TAILSCALE_IP="$1"
+SERVER_ADDRESS="$1"
 
-if [ -z "$TAILSCALE_IP" ]; then
-    echo "Usage: $0 <tailscale-ip>"
+if [ -z "$SERVER_ADDRESS" ]; then
+    echo "Usage: $0 <ip-or-domain>"
     echo "Example: $0 100.83.66.30"
+    echo "Example: $0 kraken.tail168755.ts.net"
     exit 1
 fi
 
-# Validate IP format
-if ! echo "$TAILSCALE_IP" | grep -E '^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$' > /dev/null; then
-    echo "Error: Invalid IP format. Expected format: xxx.xxx.xxx.xxx"
-    exit 1
+# Detect if it's an IP address or domain name
+if echo "$SERVER_ADDRESS" | grep -E '^[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}$' > /dev/null; then
+    IS_IP=true
+    echo "Generating SSL certificate for localhost and IP: $SERVER_ADDRESS"
+else
+    IS_IP=false
+    echo "Generating SSL certificate for localhost and domain: $SERVER_ADDRESS"
 fi
 
-echo "�� Generating SSL certificate for localhost and $TAILSCALE_IP"
-
 # Determine the output directory - we should be in backends/advanced when running
 SSL_DIR="ssl"
 if [ -d "$SSL_DIR" ]; then
@@ -53,9 +55,15 @@ subjectAltName = @alt_names
 DNS.1 = localhost
 DNS.2 = *.localhost
 IP.1 = 127.0.0.1
-IP.2 = $TAILSCALE_IP
 EOF
 
+# Add custom address as either IP or DNS
+if [ "$IS_IP" = true ]; then
+    echo "IP.2 = $SERVER_ADDRESS" >> $OUTPUT_DIR/server.conf
+else
+    echo "DNS.3 = $SERVER_ADDRESS" >> $OUTPUT_DIR/server.conf
+fi
+
 # Generate private key
 openssl genrsa -out $OUTPUT_DIR/server.key 2048
 
@@ -72,10 +80,10 @@ rm $OUTPUT_DIR/server.csr $OUTPUT_DIR/server.conf
 chmod 600 $OUTPUT_DIR/server.key
 chmod 644 $OUTPUT_DIR/server.crt
 
-echo "✅ SSL certificate generated successfully"
+echo "SSL certificate generated successfully"
 echo "   - Certificate: $OUTPUT_DIR/server.crt"
 echo "   - Private key: $OUTPUT_DIR/server.key"
-echo "   - Valid for: localhost, *.localhost, 127.0.0.1, $TAILSCALE_IP"
+echo "   - Valid for: localhost, *.localhost, 127.0.0.1, $SERVER_ADDRESS"
 echo ""
 echo "Certificate Details:"
 openssl x509 -in $OUTPUT_DIR/server.crt -text -noout | grep -A 1 "Subject Alternative Name" || echo "   (Certificate generated successfully)"
diff --git a/start.sh b/start.sh
new file mode 100755
index 00000000..44ba6f2c
--- /dev/null
+++ b/start.sh
@@ -0,0 +1 @@
+uv run --with-requirements setup-requirements.txt python services.py start --all --build
diff --git a/wizard.py b/wizard.py
index 4d37fadd..cfc5b861 100755
--- a/wizard.py
+++ b/wizard.py
@@ -26,6 +26,31 @@ def read_env_value(env_file_path, key):
     # get_key returns None if key doesn't exist or value is empty
     return value if value else None
 
+def is_placeholder(value, *placeholder_variants):
+    """
+    Check if a value is a placeholder by normalizing both the value and placeholders.
+    Treats 'your-key-here' and 'your_key_here' as equivalent.
+
+    Args:
+        value: The value to check
+        placeholder_variants: One or more placeholder strings to check against
+
+    Returns:
+        True if value matches any placeholder variant (after normalization)
+    """
+    if not value:
+        return True
+
+    # Normalize by replacing hyphens with underscores
+    normalized_value = value.replace('-', '_').lower()
+
+    for placeholder in placeholder_variants:
+        normalized_placeholder = placeholder.replace('-', '_').lower()
+        if normalized_value == normalized_placeholder:
+            return True
+
+    return False
+
 SERVICES = {
     'backend': {
         'advanced': {
@@ -154,26 +179,30 @@ def run_service_setup(service_name, selected_services, https_enabled=False, serv
             # Pass Deepgram API key from backend if available
             backend_env_path = 'backends/advanced/.env'
             deepgram_key = read_env_value(backend_env_path, 'DEEPGRAM_API_KEY')
-            if deepgram_key and deepgram_key != 'your_deepgram_api_key_here':
+            if deepgram_key and not is_placeholder(deepgram_key, 'your_deepgram_api_key_here', 'your-deepgram-api-key-here'):
                 cmd.extend(['--deepgram-api-key', deepgram_key])
+                console.print("[blue][INFO][/blue] Found existing DEEPGRAM_API_KEY from backend config, reusing")
 
             # Pass HF Token from existing speaker recognition .env if available
             speaker_env_path = 'extras/speaker-recognition/.env'
             hf_token = read_env_value(speaker_env_path, 'HF_TOKEN')
-            if hf_token and hf_token != 'your_huggingface_token_here':
+            if hf_token and not is_placeholder(hf_token, 'your_huggingface_token_here', 'your-huggingface-token-here'):
                 cmd.extend(['--hf-token', hf_token])
+                console.print("[blue][INFO][/blue] Found existing HF_TOKEN, reusing")
 
             # Pass compute mode from existing .env if available
             compute_mode = read_env_value(speaker_env_path, 'COMPUTE_MODE')
             if compute_mode in ['cpu', 'gpu']:
                 cmd.extend(['--compute-mode', compute_mode])
+                console.print(f"[blue][INFO][/blue] Found existing COMPUTE_MODE ({compute_mode}), reusing")
         
         # For openmemory-mcp, try to pass OpenAI API key from backend if available
         if service_name == 'openmemory-mcp':
             backend_env_path = 'backends/advanced/.env'
             openai_key = read_env_value(backend_env_path, 'OPENAI_API_KEY')
-            if openai_key and openai_key != 'your-openai-key-here':
+            if openai_key and not is_placeholder(openai_key, 'your_openai_api_key_here', 'your-openai-api-key-here', 'your_openai_key_here', 'your-openai-key-here'):
                 cmd.extend(['--openai-api-key', openai_key])
+                console.print("[blue][INFO][/blue] Found existing OPENAI_API_KEY from backend config, reusing")
     
     console.print(f"\n🔧 [bold]Setting up {service_name}...[/bold]")
     
diff --git a/wizard.sh b/wizard.sh
new file mode 100755
index 00000000..02942349
--- /dev/null
+++ b/wizard.sh
@@ -0,0 +1 @@
+uv run --with-requirements setup-requirements.txt wizard.py